Data backup for a computer system's stored data involves storing, at different times, copies of the data in a secondary storage location, e.g., on a secondary storage media or device or in a secondary location on a primary storage media or device. The copies are often compressed and managed by a backup software system. Managing the copies of the computer system's stored data includes understanding the data's relevance with respect to time; some data is relevant for long periods of time (i.e. medical records or financial records), and some data is relevant for short periods of time (i.e. email server or cached data from web browsers). Managing the copies of the computer system's stored data also includes understanding the requirements for availability of the copies; some data may be required to be available with a few hours notice, some data may be required to be available with a week's notice, and some data may be required to be available despite the destruction of the source computer system. Current approaches are inefficient because they do not take into account the varied time relevance and availability requirements different data have. Keeping all data for the longest retention period applicable to any portion thereof results in unneeded data being retained, increasing storage and management costs. Likewise, keeping all data at the highest level of availability applicable to any portion thereof results in data that could not be sent offsite, e.g., being kept close at hand, again increasing storage and administrative costs. Segregating data having different retention and/or availability requirements may not always be practicable and has higher costs associated with it as well. There is a need therefore for backup management capabilities that take into account the varied requirements of data in backup data sets.